Ground Truth Stixel Dataset

We have annotated twelve stereo highway sequences with ground truth regarding the free space (stixels), some with heavy rain. Find attached the raw image data (rectified pgms, 12bit/px), the ground truth stixels in xml format, the vehicle data (velocity, yaw rate, and timestamp) and the camera geometry along with a description how to use the data. All is offered to download in two zip-files.

The ground truth was acquired with manual annotations and exploiting known ego-motion for static structures. For moving objects, a vehicle tracker [11] was used with supervision.

If you use the data, please cite the following paper:

[1] David Pfeiffer, Stefan K. Gehrig, Nicolai Schneider: "Exploiting the Power of Stereo Confidences", Proceedings of the CVPR 2013, Portland, OR, USA [download].
 
(This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.)

Stixel Ground Truth Data Set part 1 (good weather): goodWeather.zip (1.6GB)
example image: blue - ground truth stixels, red to green - stixel results as described in [1]


Stixel Ground Truth Data Set part 2 (bad weather): badWeather.zip (1.5GB)
example image: blue - ground truth stixels, red to green - stixel results as described in [1]

The provided meta data in xml has the following units:
Velocity/speed : in m/s
Timestamp: in ns
Recording framerate: 25Hz


If you are interested in more detailed ground truth stereo, that covers most of the image, we recommend
the stereo benchmark (and raw data) of the KITTI homepage:
http://www.cvlibs.net/datasets/kitti/eval_stereo_flow.php?benchmark=stereo.


For questions on how to interpret the data: mail stefan.gehrig (at) daimler.com

Ċ
Stefan Gehrig,
Apr 22, 2013, 5:03 AM
Comments